Business school employability : to what extent do Social Network sites affect self-perceived employability through social capital?

(1)

Business School Employability

To what extent do Social Network Sites Affect Self-Perceived

Employability through Social Capital?

Linde Geerinck 10417818

29/05/15 2014/2015

Supervisor: Mw. S. Pajic Semester 2, Block 3

Second Supervisor: Renske van Geffen Economics and Business

(2)

2

Statement of Originality

This document is written by Student Linde Geerinck who declares to take full responsibility for the contents of this document.

I declare that the text and the work presented in this document is original and that no sources other than those mentioned in the text and its references have been used in creating it.

The Faculty of Economics and Business is responsible solely for the supervision of completion of the work, not for the contents.

(3)

3

Abstract

In an ever-changing turbulent labor market, students benefit from new ways to facilitate their chances of becoming employed. Business students are especially concerned with job insecurity and flexibility. Social networking sites are improving the way students manage their networks and relationships, which eases their self-perceived employability. This study looks at the mechanisms involved with online networking, specifically with Facebook and LinkedIn. Their theoretical and empirical connections to the Social Capital theory provide insight into what affects perceived employability. In order to examine this, a survey among business students at the University of Amsterdam (N=115) investigates the influence of student’s social network use on their self-perceived employability, with the mediating variable of bridging and bonding social capital. The results show that social capital is indeed a mediating factor, and Facebook and LinkedIn affect bridging social capital and in turn enhance their perceptions of success on the labor market. The implications involve additions to the Social Capital theory, as well as advice for students and business school curricula.

(4)

4

1. Introduction

The economic crisis has pushed youth unemployment in Europe up to 23.6% in 2013 (“Unemployment statistics,” 2015). The labor market has become extremely turbulent and European business students have not escaped this verdict, having to individually send out on average 38 job applications before getting hired (Segdhi, 2013). Learning about the reasons behind unemployment and steps students can take to make themselves more employable contributes towards easing the gap between looking for jobs and getting hired. The analysis and advice of how business students can make themselves more ready for a job hunt is thus extremely relevant.

Self-perceived employability refers to a perception of success on the labor market (Rothwell, 2007). There is a wealth of factors that contribute towards an individual’s perception of employability; be it psychological, physical, cultural, or economic. One of these factors, an individual’s social capital, concerns one’s social network of relationships (Portes, 1998). The theory of social capital was coined by Pierre Bourdieu and involves the concept of an individual’s social network adding value to life and society (2011). In line with this theory, a network of interactions and personal relationships is built and extended through continuous interaction with others, growth of personal or professional contacts, and exchange of information. Networking can be regarded as a critical competency within the business world, especially for employment opportunities or sharing resources (Janasz, 2008). The business labor market has become insecure, enforcing job seekers more responsibility to stand out amongst their competition (Bernston, 2008). Therefore, networking, in other words improving one’s social capital, is a potential and invaluable way to enhance employability within an instable labor market.

There are many different ways a student can improve his/her network and increase social capital. This study focuses on a 21st century invention, namely online social networking, specifically Facebook and LinkedIn. Since this method of making friends and developing existing relationships with peers and potentially new employers is relatively new, it is very interesting to research its possibilities and implications in the world of perceptions of employability and job-hunting. A recent study found that 15% of US-Americans interviewed found a job through social media (“Study: Social Networks,” 2011).

There is a vast amount of research done on the skills gap between graduates and the requirements of the labor market, as well as social capital and its effect on self-perceived

(6)

6 employability. There has been, to some extent, linkages made in literature between social capital and employability within business studies (Ellison, 2007; Coughlan, Swift, Jamal, and Macredie, 2012; PsycEXTRA Dataset., 2012; Berntson, 2008). However, there has been a lack of empirical data combining online social networking, social capital, and self-perceived employability.

This study will be done within the University of Amsterdam (UvA) in order to fill this gap. Therefore, the central research question of this study is the following; To what extent do

business student’s online networking activities influence their self-perceived employability through enhancing their social capital? The sub-questions that will constitute the build-up of the

main question are;

1. How active are UvA business students on online social media (in particular Facebook and LinkedIn)?

2. How does the extent of this online activity affect how students feel connected to the people around them (social capital)?

3. How does the extent of their social capital affect the perception of their own chances for success on the labor market (Self-Perceived Employability)?

Data on all three main variables, which are online social networking, social

capital, and self-perceived employability, will be individually collected. The translation of these variables into the survey will be done along the lines of models created by experts in respective fields.

This is a deductive study because its assumptions have their origin in the well-established theory of social capital. Moreover, we aim to contribute towards the Social Capital theory through the empirical examination of its implications in the particular domains of online interaction and graduate employability. However, we would like to stress that this is a predominantly practice-driven study, because its aim is to provide an advice to students, based on the results, on how to improve employability in the work force and job-hunt guidance. If online social networks are found to be positively correlated to self-perceived employability, an advice can henceforth be provided to students on how to enhance their online presence. Furthermore, detailed analysis on usage and frequency of Facebook and LinkedIn will comprise an in-depth instruction.

The level of analysis will be within a university, namely the UvA. The conceptual model will be transformed into a data-collectible form, drawn from university students. The results of

(7)

7 the study will aim to be representative of all business students at UvA. The best viable way in which the research questions can be analyzed, will be in the form of a one-time online survey-questionnaire sent out to students of the UvA.

The structure of this paper firstly entails a comprehensive literature review, exploring the main constructs to solidify the research and set a theoretical background for the analysis. The concepts of self-perceived employability, social capital, and online social networking will be explained, before being linked together into a conceptual model which will constitute the third section of this paper. The methodology in the fourth section will present the results and analysis from the surveys taken, along with clarification of validation and trustworthiness. To finalize, this will be followed by a succeeding discussion and conclusion to summarize the research, and its real-life implications.

2. Literature Review

In order to justify the relevance of the topic, previous literature must be reviewed. Firstly, the subject of employability will be explored through the internal aspect, meaning the controllable factors of employability of a graduate of an academic business-program. Secondly, the external factors of employability will be presented, referring to the current labor market dynamics, and how this is shaping employability. This will be followed by an explanation as to why self-perceived employability is a good predictive measure for actual employability to justify the measures taken in this study. The world of social capital will then be explored, starting with Bourdieu’s theory from his book “Forms of Capital” (1986) and specifying the relevant usage of his theory for the particular case of enhancing business school employability. Finally, in order to tie all the strings together, a theoretical discussion on online social networking will be given, with a section on Facebook and LinkedIn, relating back to how it can improve social capital and in turn have an effect on self-perceived employability.

2.1 Graduate Employability in a Turbulent Labor Market

Considering the vast pool of qualified business students and supply of business positions in Europe, the business labor market is an unnerving and difficult place for graduates (Sedghi, 2013). Graduate employability broadly refers to the extent of a graduate student’s ability to achieve employment (Berntson, 2008). Employment, in the context of this research, will take the

(8)

8 meaning of acquiring a graduate job, categorized as fitting to a graduate’s level of qualification (Crocker, 2012). Employability is influenced by economic, psychological, financial, political, and situational factors. The explanations and arguments are vastly spun over various disciplines and angles. For the scope of this research, self-perceived employability will be used as an indicator of actual employability, due to the fact that the participants are students, currently still unemployed. Hence, the closest proxy of their employability is how they see it themselves. The measure of self-perceived employability is very common throughout existing empirical research on graduate employability. Katz and Kahn (1978) argue that perceived employability is even more important than real employability because it is the perception of a situation that affects one’s behavior, feelings, and thinking process.

Bernston (2008) argued that throughout the literature, graduate employability is viewed through two main angles: through external factors with the dual labor market theory and internal factors explained by human capital theory, which is an umbrella theory comprising of social capital. This refers to the uncontrollable conditions of the labor market, and the internal resources and attributes related to acquiring new employment, which are also interdependent.

The dual labor market theory, coined by Doeringer and Piore, states that labor market dynamics, governed by economic shifts, are crucial shapers of an individual’s employability (Bernston, 2008). In recent years, the global labor market has become extremely volatile with increasing demands of adaptable labor force due to globalization and rapid technological advances (Kalleberg, 2001). From 1999 to 2003 the US Bureau of Labor Statistics investigated the extent of this phenomenon through surveys on job openings, hires, quits, and layoffs (Heijden, Boon, Klink & Meijs, 2009; Sadeghi, Spletzer & Talan, 2009). They found that employees had to routinely update technical and networking skills, and spend more time looking for jobs (Heijden et al., 2009; Sadeghi et al., 2009). The Netherlands is certainly no exception to this shift; there have been countless reorganizations, and a vast amount of employees has become redundant. Studies show that staff of large enterprises has felt a transition from a culture of “life-time employment” to a feeling of job insecurity, unrest, and competitiveness (Schellekens, Paas, Verbraeck, & Merriënboer, 2003).

In 2013, EU Youth unemployment stood at 23.6%, largely due to the economic crisis, and it is clear that business graduates certainly do not escape the effects of this turbulence and instability (“Unemployment statistics,” 2015). Job-seeking business students also face a threat of

(9)

9 growing competition of highly skilled international students, alongside the youth migration wave in Europe (King & Ruiz-Gelices, 2003).

Hence, the significant implications of these volatile and erratic labor market dynamics on the workers and work seekers, signify that they themselves should to be equally flexible. Graduates must become increasingly mobile and accommodating to changing work environments. There has been a transition from life-long traditional labor agreements, to employees obtaining increasing responsibility for their own career security (Bernston, 2008). This is where having a reliable network of relationships to turn to in times of need is an invaluable asset. Sharing information about technological or industry knowledge as well as information on future job prospects is a vital response to uncertainty in the labor market.

According to Janasz (2008), networking can be seen as a “key human capital skill that is unique in its ability to increase an individual’s social capital”. Being flexible and having a strong social capital constitute the internal and controllable factors of human capital on which an individual can work in order to enhance his/her employability (Bernston, 2006).

Human capital theory, introduced by Gary Becker, is the term that encompasses an individual’s resources and social capital, and is the remaining determinant of employability. It is based on the notion that training, education, and skill acquisition are crucial personal investments for employability and career success (Bernston, 2006). Human capital is knowledge that is physically inseparable from a person, unlike financial capital.

Rothwell (2008) conducted a study in the UK with the purpose of developing a sound research instrument to measure self-perceptions of the employability of business undergraduates. While exploring the concept, he found it to be multifaceted and concluded that four main components were at play, which can be seen through the framework previously mentioned of internal and external factors. He found two components within the internal human capital spectrum that contribute towards a student’s perception of employability; the participant’s

self-belief and motivation, and field of study. He also added stated of the external market labor market

and influence of one’s university brand as third and fourth components which in turn plays with the internal factors. He then composed a matrix displaying the interplay between all four components;

(10)

10

Figure 1. Self-Perceived Employability Matrix

My University

1. My engagement with my studies and

academic performance

2. My perception of the strength of the university’s brand 3. The reputation my university has within my field of study Self-Belief 8. My confidence in my skills and abilities

4. The status and credibility of my field of study My Field of Study 7. My awareness of opportunities in the external labor market

6. My perception of the state of the

external labor market

5. The external labor market’s demand for people in my subject field

The state of the external labor market

For every cell he devised two questions for a validated survey which will be the theoretical backbone of the empirical data collection of self-perceived employability in this study.

2.3 Social Capital

There are various ways of enriching human capital, such as competence development through internships, work experience, or training (Bernston, 2006). These competences include skills and attributes such as problem solving, having an entrepreneurial mindset, or being adaptable to change, and positively correlate to employability. However, these personal skills are frequently studied. On the other hand, social capital has been studied less in the realm of boosting human capital.

Social capital enhancement is an invaluable asset in lieu or as an addition to practical experience. Research, including that of Granovetter (1973), has shown that successful job seekers tend to score higher on social capital tests. The complex theory of social capital is one that has been extensively used in the social sciences, being especially popular during the last 20 years, and is very useful to suit the study of social implications of networking (Janasz, 2008). The

(11)

11 pioneer of the term, Pierre Bourdieu (2011), refers to social capital as links and shared values in society that enable individuals to trust each other and work together. During the past two decades, Bourdieu’s theory has been vastly applied in the fields of management and organizational behavior (Carpenter, Li, & Jiang, 2012). Relating to economics, it can be seen as the power of nonmarket relationships influencing individual and collective behavior (Durlauf, 1999). Research has brought to light various positive implications on micro and macro scales. On a national level, strong social capital has been shown to improve public health, lower crime rates, and boost financial markets (Adler & Kwon, 2002). On an individual scale, apart from being linked to psychological well-being, the sharing of resources is the benefit of social capital most observed by researchers. These resources can come in the form of the capacity to mobilize group action, personal relationships, or information (i.e. about new job openings). Social capital across groups can provide access to non-redundant information, and can lead to employment connections (Granovetter, 1973).

Negative attributes have also been found to result from the forming of groups, such as discriminatory behavior towards other groups (Durlauf, 1999). Even the role of social capital in the realm of hiring has a negative side. It needs to be taken into account that the influence of a relationship, within the context of job search and personnel selection, has not been seen only through a positive glance. The dark side of social capital can include the interplay of social inequalities and uneven opportunities across employment structures (Gargiulo & Benassi, 1999). This will not be addressed in this paper because its purpose is to investigate ways of making people more employable on an individual scale, but it is interesting to keep in mind. Putnam, a respected leader in the field of social capital theory, believes in the overriding benefit of social capital in contrast to its downfalls (Putnam, 2004).

Many scholars argue that the term has been used in so many ways and for so many distinct sociological explanations that the definition of social capital has become “contentious and slippery” (Smith, 2010). Therefore it is crucial to make explicit which explanation and version of social capital will be used in the context of this research. We will adopt social capital, especially in the scope of human resources and organizational psychology, as a concept split up into two main categories; bonding and bridging. This focus lies on the breadth, depth, and nature of ties between individuals. The scope of a tie (or relationship) refers to time spent, emotional intensity, intimacy, and reciprocity (Kavanaugh, Reese, Carroll & Rosson, 2005).

(12)

12 Bonding social capital describes the trust provided by strongly tied individuals, mostly family and close friends. There is usually a minimum of diversity in terms of background of the individuals, but the personal connections are intact. Putnam argued that it reinforces exclusive identities and like-minded individuals (2004). Williams (2006) categorizes it as;

 Emotional support

 Access to scarce resources

 Ability to mobilize solidarity

 Out-group antagonism

Bridging social capital, according to Putnam (2004), occurs when individuals differing in backgrounds and communities make connections. These relationships are mostly tentative, not fixed, but can unlock the door for new opportunities, information, and resources. Weak ties have the ability to link members and groups differing in social settings, such as geographic location, or otherwise disconnected. These weak ties in a society are said to be the backbone of healthy, modern societies, helping the flow of ideas and information (Newton, 1997). Content of information could include a new job opening up, a free job position or the opportunity for a job interview. That is why bridging social capital is the most interesting in the context of this research and the topic of employability chances. ‘Weak ties’ is another term used by many scholars to describe bridging relationships. Williams (2006) broke weak ties down into four parts;

 Outward looking

 Contact with a broader range of people with different background

 Viewing oneself as part of a larger group

 Reciprocity with a broader community

This distinction between bridging and bonding social capital has brought to light the different layers of social relationships that exist and can be enhanced by networking to lead to employability. It must be kept in mind that the line between bridging and bonding social capital is often blurred and intertwined. Nevertheless, for the ease of empirical research, the roles of bridging and bonding social capital will be examined separately, each in relation to social networks, before being linked to employability.

(13)

13

2.4 Online Networking Sites to Increase Social Capital and in Turn Employability

In this study, the method of improving social capital and in turn employability concerns the building and maintaining social relationships with the use of social network sites (SNSs). Since their introduction, SNSs such as Facebook, LinkedIn, and Myspace have obtained millions of members and have become part of the everyday lives of their users (Boyd and Ellison, 2007). A Pew Research Center survey found that the number of memberships of SNSs has doubled since 2008. The purposes of using online social networking vary across users. Beyond their profiles, users can share pictures or videos, blog, or send instant messages.

What makes internet social networking so interesting, is its ability to connect people irrespective of physical location barriers. Through a community-based discourse, virtual groups can be formed, online debates can be held, and information can be shared with a large group of people. LinkedIn, for example, reaches 200 countries (“About Us,” 2015). In terms of finding employment, this platform has a huge potential simply because of mobility (Kiesler, 1997).

This virtual phenomenon has attracted the attention of many scholars in terms of its implications on the way society communicates and interacts, as well as gaining an understanding of the practices, culture, and meaning given to these virtual spaces. On one hand, Internet use “detracts from face-to-face time” with others, which could diminish the user’s social capital (Cronk and Gurteen, 2012). However, there is more evidence that internet communication leads to positive effect on sense of community, trust, and reciprocity (Chiu, Hsu & Wang, 2006). A study done by Pew Research Center in 2006 found that online users tend to have a larger perceived network of ties than non-users, and were more likely to receive help from core members in their network.

The main way the use of social network sites (SNSs) improves employability is through enabling the building and maintaining of social relationships and networks. Focusing on student population, Ellison, Steinfield, and Lampe conducted a survey-based study on undergraduates in 2007, which measured the effect of their Facebook usage on their overall social capital. This study will be the empirical support of the ‘SNS-social capital’ section testing of this paper, further elaborated on in the conceptual model. In Ellison et al.’s research, their aspect of social networking usage was divided into three characteristics. Once it was known that a participant was a member of a site, in their case Facebook, they argued that the intensity of usage was important, distinguishing between active engagement, the extent of emotional connection towards the SNS,

(14)

14 and the extent of integration in daily life of the network (Ellison et al., 2007). Their scales are credible and valid, which is proven by the fact that they were used in many proceeding studies by other researchers without much alteration, providing useful insights for these studies (Sgambato et al., 2011; Coughlan et al., 2012; Benson 2013). Students are especially interesting to study since in the context of international SNS use university students and young people are “largely early adopter and heavy SNS users” (Ellison et al., 2007).

Facebook and LinkedIn were chosen as representatives of SNSs for this research, following the study of Coughlan et al. in 2012. We chose to focus on these two social network sites for several reasons. Facebook is one of the most-visited websites in the world and currently boasts more than half a billion users (webneel.com, 2015). Originating at Harvard, Facebook has become an integral part of the majority of university students lives in Europe. Facebook is widely known in the Netherlands, and is therefore an optimal choice of online social network for this study. The network is especially useful to study due to its capacity to bring online relationships into the offline community, which in this case relates to social capital.

With over 150 million members, LinkedIn is purely for the purpose of making business connections. It is the leading professional online social network and therefore extremely relevant when analyzing one’s network in terms of employability (Ellison, 2007). It is estimated that 94% of recruiters in the US use LinkedIn to scout new talent (Fernando, 2014). Due to the success of the website in the US, it will be interesting to see if students at the University of Amsterdam have adapted the practice. If not, this could lead the section in this paper on further possibilities for studying online interaction in the Netherlands.

Studies on competence development, such as problem solving skill acquisition, has been on the forefront of human capital research. Less investigation has been done on online networking that enhances social and human capital. One reason for this is because the Internet and hence SNSs are not older than 30 years and so scientists have not yet fully incorporated the phenomenon into solid theories. Researchers who study employability are often more focused on the academic content of business-programs, rather than examining employability as a result of interpersonal skill and networking. This study aims to fill the gap of research surrounding online networking.

(15)

15

2.5 Conclusion

The relevance of social networking in terms of improving a student’s social capital and hence his/her self-perceived employability has been presented. It is clear that the aspect of employability is determined by a wealth of factors and thus multi-faceted and complex.. Through defining and explaining concepts of employability and self-perceived employability, parameters and scope of this research were established.

The current state of the volatile labor market shows the need for analysis and contemplation on how students can keep themselves attractive amongst growing competition. Advice on how to stay flexible can be supported by the theory of human capital. in order to investigate each factor in depth, they need to be looked into independently. To stay realistic for the length of this study, only one factor was chosen, social capital. This factor,and more specifically the potential of online networking to increase social capital, was chosen due to a lack of a pre-existing body of academic work. Therefore, the impact of SNS usage on social capital and in turn on self-perceived employability will be measured amongst business students at the UvA. An in-depth look into how this theory conceptualizes and is moderated empirically will be given in the following sections.

3. Conceptual Model

After having explored the previous literature surrounding the three main concepts that build the picture of this study, the following conceptual framework will extend the literature into a set of hypotheses that will guide the data collection as well as the analysis. The first piece of the puzzle will constitute exploring the effect of SNSs on social capital. Secondly, social capital will then be tied with self-perceived employability. Finally, to check if social capital is indeed a mediator, expectations between the relationship of SNS and self-perceived employability will be investigated.

3.1 SNS and Social Capital

In line with the previous distinction of bridging and bonding social capital, internet or SNS usage also has varying different effects on both strengths of relationships.

In terms of bonding social capital, Ellison et al. found that offline strong relationships could be maintained, solidified, or amplified online (2007). In interviews they conducted,

(16)

16 students gave the example of the “Birthday Reminder” feature of Facebook, which aids the maintenance of relationships.

Other researchers, however, found a lower positive correlation between bonding ties and Facebook usage than with bridging ties. In fact, the evidence of positive effect of SNSs on bonding or strong ties is much less solid throughout other literature. Coughlan et al. found a positive association with bonding and Facebook usage, but the results were not significant enough to accept (2012). Sgambato (2012) also used both Ellison et al.’s scales and hypothesized positive implications of Facebook intensity on bonding social capital, which wasn’t supported and the hypothesis was rejected. He argued that a reason for this could be that deep and emotional ties are not mainly tended to offline and not directly affected by SNS usage (2012). That is why in this study, the prediction on the relationship between SNS usage and bridging social capital and bridging ties will remain ambiguous until the results section, providing room for deeper analysis. Therefore the first hypothesis, keeping in mind our distinction between Facebook and LinkedIn will be;

H1a: The Intensity of Facebook usage will affect the student’s bonding social capital. H1b: The Intensity of LinkedIn usage will affect the student’s bonding social capital.

Various authors have made links between SNSs and the forming of bridging social capital, or weak ties (Ellison, 2007). Forming new connections or networks online can be possible through using processes like photo directories, name lists, or search engines (Resnick, 2001). The Internet can lower barriers to participation due to its easy, mobile, and cheap nature (Coughlan, 2011). In this sense, the infrastructure of technology is well suited to maintain or initiate bridging relationships. In his study of online vs. offline social capital, Williams uses the term ‘online bridging,’ and found that his measures showed examples of the concept; coming into contact with unlike people, links to information outside one’s daily routine, and behavior such as meeting new people in chat rooms (2006). Hence, the next hypotheses deduce to;

H1c: The Intensity of Facebook usage will be positively associated with the student’s bridging social capital.

(17)

17

H1d: The Intensity of LinkedIn usage will be positively associated with the student’s bridging social capital

3.2 Social Cap and Self-Perceived Employability

As shown in the literature review, there is a lot of theoretical evidence that social capital is a part of employability. Therefore a boost in social capital, be it bridging or bonding, would lead to more chances on the labor market. However, there is a lack of empirical evidence using students which supports this claim. Mancinelli did a study in 2010 which showed that Italian immigrants working in Europe relied on existing ethnic networks for employability. There are also many studies done on the effect of network on career success; (Cannings, 1988; Gould & Penley, 1984; Peluchette, 1993). Career success is relevant because the perception of to what extent a student feels fit to be employed is a predictor of career success, as shown by Wittekind, Raeder, and Grote in 2010.

A similar comparison can be made between bonding/SNS usage and bonding/perceived employability. Throughout the literature, the relationship between bonding and self-perceived employability is not as directly correlated. This is because a future employer or employment connection usually happens amongst a student’s weak ties, as found by Granovetter (1973). Therefore, the influence that strong bonding social capital exerts on employability works through affecting social behaviour and identity, and hence intensifying expectations of employment. This association is much more abstract than the bridging aspects with SNS and employability. However, the two factors cannot be isolated because they affect each other. That is why it will be hypothesized that;

H2a: Bonding social capital will be positively associated with the student’s self-perceived employability.

Weak ties present a clearer link with employability. The main reason for this is the access to resources and information outside one’s immediate circle, which Granovetter stated is optimal for hiring opportunities (2000). As stated beforehand, Putnam’s components of bridging social capital are; outward looking, contact with a broader range of people, a view oneself as part of a broader group, and diffuse reciprocity with a broader community. The emphasis on a network comprised of people with dissimilar demographics validates Granovetter’s claim. The broader

(18)

18 range of backgrounds, when consolidated by reciprocity, will provide access to information that would otherwise not be available. He showed this by asking labourers in professional, technical, and managerial jobs if they found their job through contacts. Upon their agreeing answers, he asked them how well they know the person, establishing the strength of the tie. His results showed the relationship between the hired and the job provider to be mostly “sporadic” (Granovetter, 1973). This leads up to our next hypothesis;

H2b: Bridging social capital will be positively associated with the student’s self-perceived employability.

3.3 Social Capital as a Mediator

Lastly, the literature review has shown that evidence of SNS usage contributing towards enhancing a student’s self-perceived employability is always through the mediating role of social capital. The direct links made in the literature between SNS and employability exempting social capital mostly surround the concept of ‘online recruiting’ (Joos, 2008). These include discussions about recruiters going online to find employees and the shift from traditional newspaper ads to online ads (Barras, 2006).

This realm of study is very new, and the studies are mostly informational about the implications of internet use in general, of which recruiting has become a growing section of. However less scientific, this perspective is also interesting, but takes different pathways to this study. That line of thinking does not contradict this one, because the SNS usages investigated in those studies are purely about job searching and employee recruiting, and not about social relationships and community feeling in general. Therefore, social capital in this study is a fully mediating variable. We expect that the contribution to social capital building will fully explain the positive influence of SNS usage on graduate employability. Specifically, we expect that the usage of SNS will reinforce the development of both strong and weak ties, which will consequently positively affect perception of employment prospects, defined through the following hypotheses;

H3a: The effect of Facebook usage on self-perceived Employability is fully mediated by social capital.

(19)

19

H3b: The effect of LinkedIn usage on self-perceived Employability is fully mediated by social capital.

These hypotheses have put conceptual arches around the theories discussed, as well as conceptualized the relationships between them, in order to ensure sound a basis for the empirical data of this study.

4. Methodology

This section will show how the central research question was operationalized and empirically carried out in the form of a survey. The questions used for the survey were taken directly from existing literature and produced measures for SNS usage, Social Capital, and Self-Perceived employability. The questions were then slightly altered to be relevant in the context of this study. The design of the research will be explained, followed by a description and parameters of the sample. The section ends with the measures used, which all includes limitations of the methodology.

4.1 Research Design

The research method used for this study was an online, self-administered, questionnaire-based survey, which measured each variable through independent sets of questions. There are several reasons a survey was optimal for this quantitative study with the purpose to find correlations between three determinants and social constructs (SNS usage, social capital, and self-perceived employability).

Firstly, the method fits the deductive nature of this study. The initial interest to conduct the research was a desire to explore the relationship of two elaborate and thorough theories of social capital and employability. This compelled the deductive approach. Furthermore, Bryman and Bell (2007) argue that deductive research is more organized, has a higher degree of certainty than an inductive approach, and follows a sequence formed by logical reasoning. A mostly closed-question survey can be formulated by previous theories, which will benefit from a sound and validated basis through which complex analysis can be conducted. In addition, Saunders (2009) argued that people are more hesitant to answer open questions. A limitation, however, to a mostly closed-question survey is the rigidness of the answers (Saunders, 2009). Participants are to choose an answer from a limited list of responses, without being able to provide their own

(20)

20 ideas. Because this is not an exploratory study, and the theories are being tested, this drawback is not so threatening.

Another reason for using the survey method is that, contrasting to interviews, surveys are efficient, low cost, and can reach a large amount of individuals (Saunders, 2009). The efficiency aspect is very attractive in this case due to a limitation of time. A large sample size is particularly valuable for generalizability (Saunders, 2009). This research strives to gain understanding of the characteristics of business students in the Netherlands in order to provide implications for students and job-seekers. In this case, the participants are only from one university in Amsterdam, but a large sample size will allow more generalizability than other methods using less participants. Another benefit of a large sample and survey structure is that the standardized and consistent questions will enable vast comparison potential among the individuals.

The hypotheses in this research have been replicated directly from other researchers (Coughlan et al., 2012; Ellison, 2007; Sgambato, 2012). This allows for a direct comparison between the results of this study and the results of the hypotheses of the discussed authors. In light of different results, dissimilarities between the characteristics can then be approached for further explanation.

4.2 Research Sample

This research sample aims to provide representation for the population of university business students studying in the Netherlands. The students in the sample are all from the University of Amsterdam (UvA) because they were the easiest to approach. The fact that only students from the UvA in the Netherlands were participating is a limitation to the generalizability of the data. Certain characteristics could stem purely from attending this particular university. This becomes important when looking at the specific scales of the survey. Among the dimensions of Self-Perceived Employability, Rothwell mentioned influence of university brand as having an effect on the measure (2008). Therefore, students from different universities in the Netherlands might score differently on the self-perceived employability test because of their university. However, this limitation has to exist in order to optimize the sample size.

In order for the sample to be representative, variations of demographics such as age, nationality, and gender should strive to have the same build-up as its population. This was taken into consideration. After the population was established, we approached the participants in the form of convenience sampling, a widely used method (Saunders, 2009). This involved selecting

(21)

21 the cases simply because they were available, convenient, and willing to fill in the survey. Even though this method is prone to bias and influences beyond our control, these samples often meet relevant purposive sampling criteria. In addition, it might not be so problematic in this case because in general, the population of business students are relatively homogenous in terms of age and class (Henrich, Heine, & Norenzayan, 2010).

4.3 Data Collection

Due to the survey being web-based, the question were assembled using the website www.qualtrics.com, and distributed by means of a url-link. This website was very user-friendly and saved time because of its pre-existing proxy demographics questions. The creation of the survey was done in collaboration with three others students researching for their thesis. They were all investigating the effect of different independent variables on self-perceived employability. These variables consisted of: entrepreneurial mindset, internship experience, and personality attributes.

The questionnaire was a combination of all variables (including SNS usage and social capital). The benefits of this method included a potential to efficiently enlarge the sample size because four researchers would be actively looking for participants. However, a limitation to the collaboration was the inevitable long size of the survey, which ended up being around 20-25 minutes. Saunders (2009) argued that a lengthy survey can prohibit participants to complete it or to start it in the first place. Our survey was lengthy because everyone’s variables had to be accounted for. This had an adverse effect on the respondent rate because some people did not have the patience to finish it. However, we consider this as a trade-off for working together.

The participants were then approached through Facebook personal message, E-mail, or face-to-face based on the researcher’s personal networks. It was also posted on the university’s Facebook page as well as the specific ‘UvA Business Students’ page with 53 members. The fact that students were approached through online social media will definitely have an effect on the way they answer the SNS questions. These participants were obviously members of the network sites. However, those questions measure the frequency and intensity of the online usage, and variations amongst participants will still be possible.

Another constraint of web-based surveys is the apparition of self-selection bias. According to Lavrakas (2008), this occurs when respondents can decide whether they want to participate in the survey or not and can be detrimental to the internal validity of a study. He

(22)

22 argued that a problem could be that the respondents are then not representative of the population because their propensity to participate is correlated with interest in the topic being researched (Lavrakas, 2008). This was slightly eradicated by providing a monetary incentive to participate, which meant that other reasons to complete the survey were present.

The survey was piloted by ourselves, by which we corrected the found errors and redundancy. In addition, the section of social capital received some alterations with the goal of shortening the whole survey length. The final survey was estimated to take around 20 minutes.

4.4 Measures

The attributional and behavioral measures used in the survey were made to find indications of each participants SNS usage and intensity, their bonding and bridging social capital, and their self-perceived employability. All scales were existing from the literature, and have already been tested and validated, increasing the validity of this study substantially. The survey is fully shown in Appendix A.

4.4.1 Demographics and Control Variables

Firstly, we collected information on the demographics, which included gender, nationality, and level of education (bachelor or master). These variables all constitute our control variables. Control variables are in addition to the independent, dependent, and mediating variables, and should not affect the outcome. They are still tested to help rule out alternative explanations of the relationships (Saunders, 2009).

Gender was used as a control variable in all the studies that the scales are based on (Ellison et al., 2007; Rothwell and Arnold, 2008), despite not finding significant differences. The sample of this study shows a considerable higher percentage of females (66.1%), which is not close to the actual percentage in the population. In 2014, the course Economics and Business bachelor contained 66% male and 34% female (“Student Enrollment”, 2015). However, Elison et al.’s (2007) study also had 66% females and did not find any sign of correlation between gender and the other factors. Level of education is relevant to see if students that have been exposed to more university levels will have significantly different perceptions of their own employability. The nationality indicator was divided into two categories: Dutch and Non-Dutch. This last variable will make it possible to see if the fact of living abroad plays a role in one’s social capital.

(23)

23

4.4.2 Facebook and LinkedIn Usage

SNS usage is our independent variable, and comprised the first variable measurement after the demographics section of the survey. The questions were completely replicated from Ellison et al.’s 7-item scale for business students in the UK. However, their study only measured Facebook usage, and so we adapted the questions for LinkedIn as well. The Cronbach’s alpha, showing the validity for their scale, is 0.83. They distinguished Facebook usage between three categories, and measured each accordingly;

Table 1. Facebook Intensity Scale

Component Question Response Style

active engagement - how many Facebook “friends” do you

have?

- how many minutes do you spend on Facebook a day?

Open answer

emotional connection - I feel out of touch when I haven’t logged

onto Facebook for a while

- I feel I am part of the Facebook community

- I would be sorry if Facebook shut down

5- Point Likert Scale

integration into daily life - Facebook is part of my everyday activity

- Facebook has become part of my daily routine

5- Point Likert Scale

The LinkedIn section looked exactly the same as the Facebook questions, except for the questions about integration into daily life being different. Due to its purpose, LinkedIn is not a website that participants use daily. It is mostly used for solely contacting bridging ties to make professional links (Coughlan, 2012). Therefore, instead of using the statement “LinkedIn has become part of my daily routine,” the question was changed to “LinkedIn has become part of my

weekly routine.” Also the question ‘Facebook is part of my everyday activity’ has transformed

into ‘LinkedIn is an important part of my career pursuit,’ to make it more suitable.

4.4.3 Social Capital

Social capital as a mediating variable was the second variable asked in the survey. The reason for this was the hope that participants would be in the line of thinking of their personal network after receiving the questions about Facebook and LinkedIn. The questions for social capital were taken from Ellison et al. (2007) with a Cronbach alpha of 0.87. The scale was also conceptualized from

(24)

24 Williams’ scales (2006) who designed a scale inspired by Putnam’s theories on social capital (2004).

Table 2. Social Capital Scales

Bridging Social Capital - I feel I am part of my community - I am interested in what goes on around me - I come into contact with new people all the time

- The people I interact with would be good job references for me

Bonding Social Capital - There are several people I trust to solve my problems (eg: lending me an emergency 100 euros)

- There is someone I can turn to for advice about making very important decisions - I do not know people well enough to get them to do anything important (eg:

helping me move)

Many researchers from varying disciplines have used this scale when analyzing social capital, specifically in the realm of the online community (Kim & Kim, 2014; Coughlan, 2012). Williams (2006) was the first to design the scales, and his purpose was to find differences between the social characteristics of the online versus the offline sphere. Therefore this scale is very adequate in the context of this research. A big limitation is that while the survey was being shortened for efficiency, a lot of questions from this section were taken out. Having only 4 questions for bridging and 3 for bonding capital will make the data very difficult to be reliable.

4.4.4 Self-Perceived Employability

This variable, all four researcher’s dependent variable, was taken from Rothwell and Arnold’s scales from 2008 with a Cronbach alpha of 0.75, which were in turn adjusted from employee to student self-perceived employability from Rothwell and Arnold (2007). The scale comprised of 11 statements, which again had 5-point Likert-scale responses. The four components introduced by Rothwell in 2008 each correspond to separate statements in the survey;

Table 3. Self-Perceived Employability scales

Self-Belief and Motivation - I achieve high grades in relation to my studies - I regard my current studies as top priority.

- I can easily find out about opportunities in my chosen field.

- The skills and abilities that I possess are what employers are looking for. - I am generally confident of success in job interviews and selection events. University Brand - My business school has an outstanding reputation in my field of study.

(25)

25

Field of Study - My future profession ranks highly in terms of social status.

- People in the career I am aiming for are in high demand in the labor market. External Labor Market - There is generally a strong demand for graduates at the present time.

- There are plenty of job vacancies in the geographical area where I am looking.

A limitation of using this scale could in fact be the inclusion of the ‘external labor market’ conditions. While this is composed into an indicator of self-perceived employability, it is a factor that the respondent has no control over. Hence, an increased social capital will not have a direct effect on how the respondent perceives the labor market. However, due to self-perceived employability being multifaceted and dependent on those factors, they cannot be left out of the full measure.

4.5 Methodology Limitations

One limitation of the study could be the apparition of participant fatigue. This occurs when participants feel the survey task is taking too long, become tired, and the quality and focus of their answers starts to deteriorate (Lavrakas, 2008). After piloting our survey, we found it to take around 20 minutes. This is still a reasonable amount of time, but borders the maximum time that is appropriate. Additionally, we found a lot of cases where respondents started the survey but failed to complete it. Participant fatigue could have been the cause of this.

Another limitation is that due to the survey being shared mainly through Facebook, automatically all the respondents were Facebook members. This left no potential to compare social capital and employability between users and non-users. Figures showed that in 2012, 86% of UvA students had a Facebook profile. Unfortunately, an effort was made to seek non-members but none were found. It is possible that the percentage of members is higher in 2015 and that non-members are rare.

5. Results

After discussing the layout and dimensions of the survey in the methodology, this next section will display the results of the data analysis with the software SPSS. Firstly, an overview of the sample characteristics in the form of descriptive statistics will be given, followed by a reliability test with calculating Cronbach’s alpha for each main variable. Subsequently, a correlations table

(26)

26 will be shown to provide initial analysis, followed by the results of multiple regressions and two mediation tests.

5.1 Descriptive Statistics

Firstly, the data was scanned for incomplete surveys. From the 133 respondents, 18 were removed due to filling in less than 5%, leaving N=115 for the final sample size. However, amongst these 155 surveys, some responses were carelessly filled in or some answers were left out. These faults were picked up by SPSS and left out through listwise deletion per individual analysis, depending on how much data of each participant could be salvaged for each test. The mean and standard deviation of all the variables are to be found in the table in Appendix B. The following table presents the demographics of the sample.

Table 4. Demographics and Control Variable Statistics

N=115 Percentage (n) Gender Male 33.9% (39) Female 66.1% (76) Level of Education Bachelor 69.6% (80) Graduated from Bachelor 11.3% (13) Master 19.1% (22) Nationality Dutch 39.1% (70) Non-Dutch 39.1% (45)

5.2 Reliabilities and Correlations

A reliability test is firstly executed in order to determine whether data analysis was plausible. This is accomplished with the Cronbach’s alpha test, which is the most common measure of internal consistency (Saunders, 2009). All the scales were checked to see if any items should be deleted, which was not the case. The Cronbach’s alphas are shown on the diagonal in the table in Appendix B. It is noticeable that while the Cronbach’s alpha for Self-Perceived Employability (.82) is good and Facebook Intensity is reasonable (.73), the rest are below 0.7 (LinkedIn Connections (.55), Bridging (.57), Bonding (.66), Social Capital (.66)). In the social sciences, a minimum of 0.7 is recommended (Saunders, 2009). These considerably low alphas are a

(27)

27 limitation to this study. The LinkedIn Intensity measure gave a very low Cronbach’s alpha (.55), and did not show any significant correlations to social capital or self-perceived employability. Therefore, it was decided that the indicator for LinkedIn usage would instead be the number of LinkedIn connections.

The correlations table also shows the correlations between the three main variables as well as the three control variables. In addition, the Social Capital variable was fragmented into Bridging and Bonding scales, with 4 and 3 items respectively. Following the predictions, Social Capital is seen to be significantly correlated (although quite low) to Facebook Intensity (r=.205), LinkedIn Connections (r=.358), and Self-Perceived Employability (r=.286). However, once separated into its subscales, bonding did not correlate with any other main variables. Bridging, nevertheless, which was seen as the more interesting type of social capital in relation to employability, did have significant relationships with Facebook Intensity (r=.209), LinkedIn Connections (r=.358), and Self-perceived Employability (r=.293). In terms of the control variables, Gender and Nationality did not appear to have any effect on the factors. Interestingly, there is a correlation between Education Level (whether the participant is an undergraduate or a postgraduate) and LinkedIn connections (r=.19), Bonding social capital (r=.277), and Social Capital (r=.773).

5.3 Regression Analysis for SNS Usage and Social Capital (H1a, H1b, H1c, H1d)

To test the hypotheses 1a, 1b, 1c, and 1d, linear regressions were run using Facebook Intensity, LinkedIn Connections, Bridging, and Bonding. H1a and H1c predicted the direct effect of Facebook Intensity and LinkedIn Intensity on Bonding social capital, and H1b and H1d tested the direct effect of the two network usages on Bridging social capital. Facebook Intensity and LinkedIn Connections were independent variables and Bridging and Bonding were used as dependent variables. In order to test the aforementioned hypotheses, we performed two independent hierarchical multiple regressions, each of them comprising of two different steps. In the Step 1 of both hierarchical linear regressions the control variable Education level was introduced. Subsequently, in order to independently test the effect of Facebook usage and LinkedIn Usage, after controlling for Education level, we introduced these two variables in the in Steps 2a and 2b of each dependent variable testing.

(28)

28

5.3.1 Bonding

The first step in the model in Table 5 was statistically significant (F(df,113) = 9.301; p< .01) and explained 37.4% of bonding, indicating that educational level exerts influence on bonding regardless of the usage of social network sites. Moreover, the positive sign indicates that students with higher educational level (postgraduates) report bonding relationships in higher extent.

Hypothesis 1a that predicted an effect of Facebook usage on bonding social capital was tested in Step 2. This analysis demonstrates the individual contribution of Facebook Intensity and Bonding social capital, after controlling for Education level, and displays addition of 1.9% (F (df,

113) = 5.919; p < .01) of explained variance, which is not significant (b=.125, p=.215). Even in the context of intensity of Facebook usage education level had positive influence on bonding, indicating that it is an important factor to control for. Even though the initial hypothesis of Facebook on bonding did not predict a necessarily positive correlation, there was no significant correlation at all. Hence, H1a is not accepted.

The next hypothesis (H1b) of LinkedIn having an effect on bonding social capital is also not supported in the model. The added contribution of the LinkedIn to bonding, after controlling for educational level, was insignificant 4.5% (F (df, 115) = 4.636; p < .05) with insignificant correlation (b=.000, p=.823). H1b is consequently not accepted. Hence, it can be said that in this context, LinkedIn Connections did not explain changes in Bonding social capital.

Table 5 contains the results of three steps or three models run. Model 1 includes the control variable Education Level (undergraduate or postgraduate), whereas Model 2 and 3 test hypotheses H1a and H1b respectively.

(29)

29

Table 5. Regression results: The effects of Facebook and Linkedin usage on Bonding, after controlling for Education Level Model 1, 2 and 3: Bonding

B SE Beta Sum of squares Df F Sig Step 1 3.392 113 9.301 Constant 3.757 .17 .000** Education Level .374 .123 .277 .003** Step 2a 4.263 113 5.919 .004** Education Level .39 .122 2.89 .002** Facebook Intensity .125 .081 .141 .215 Step 2b 3.410 113 4.636 .012* Education Level .369 .125 .273 .004** Linkedin Connections .000 .000 .021 .823 R2 .096 .122 R2 Change .019 .045 Note. N=115, *p <.05,**p<.01. 5.3.2 Bridging

The second set of hypothesis focused on the role of social network usage on bridging. As predicted through the literature review, bridging social capital gave more significant results. Moreover, we controlled again for the role of Education level, by inserting it in the Step 1 of hierarchical linear regression. However, the effect of the control variable on bridging was not as crucial as on bonding. Independently of Facebook intensity and LinkedIn connections, higher level of education did not significantly affect bridging 1.5% (F (df, 115) = 1.759; p=1.87).

Additionally, in order to independently test the effect of Facebook usage and Linked Usage, after controlling for Education level, we introduced these two variables in the model through Steps 2 and 3.

Step 2a in Table 6 shows the effect of Facebook Intensity on Bridging social capital. The results showed that 4.9% of the variances were explained by Facebook (R2_{change=.049). Even} though this value is not very high, it is still significant (b=.178, p=.018*). This means that H1c is

accepted. It can then be said that Facebook usage affects Bridging social capital after having

accounted for Education Level.

Step 2b tested the affiliation between LinkedIn Connections and bridging social capital. The explained variances were 10.7% and the results were significant (R2_{change=.107, b=.178,}

(30)

30 p=.000), hence H1d is accepted, meaning that changes in LinkedIn Connections positively affect bridging social capital.

Table 6. Regression results: The effects of Facebook and Linkedin usage on Bridging, after controlling for Education Level Model 1, 2, and 3: Bridging

b SE Beta Sum of squares Df F Sig Step 1 .562 113 1.759 .187 Constant 3.560 .159 .000** Education Level .152 .115 .124 .187 Step 2a 2.318 113 3.779 Education Level .175 .113 .143 .125 Facebook Intensity .178 .074 .220 .018* Step 2b 113 7.733 Education Level .078 .111 .064 .481 Linkedin Connections .001 .000 .332 .000** R2 .064 .122 R2 Change .049 .107 Note. N=115, *p <.05,**p<.01.

5.4 Regression Analysis for Social Capital and Self-Perceived Employability (H2a and H2b)

The next hypotheses were also tested with simple regression in Table 7, and involved the correlations between Bridging and Bonding on Self-Perceived Employability. Multiple regressions were also run to test if level of education had any effect, but no significant results were found, hence left out of the table. Bonding displayed an insignificant 2.4% variance explanation with (b=.133, p=.13). Hence, H2b was not accepted. This indicates that an increase in bonding social capital would not explain an increase in self-perceived employability. After controlling for education level, bridging, following predictions, showed significant explained variances of the employability measure and was significant (b=.291, p=.004). Hence, H2b was

accepted. Moreover, pretty high percentage of the variance in Self-Perceived employability is

(31)

31

Table 7. Regression Models with DV as Self-Perceived Employability

Model 1 and 2: Self-Perceived Employability

b SE Beta Sum of squares Df F Sig 𝑅 2 Model 1 2.476 94 8.847 .004** .293 Constant 2.351 .370 .000 Bridging .291 .098 .293 .004** Model 2 .696 94 2.33 .130 .024 Constant 2.880 .371 .000 Bonding .133 .087 .156 .13 Note. N=115, *p <.05,**p<.01.

5.5 Mediation Analysis (H3a and H3b)

Multiple regression was used for the final results section for hypotheses 3a and 3b, namely to check whether the relationships between Facebook Intensity and LinkedIn connections were fully mediated by social capital. Mediation occurs when three conditions are met. Preacher and Hayes (2008), show that in order for there to be mediation, the independent variable must have significant effect on the predicted mediator (path a), followed by the mediator significantly accounting for the dependent variable (path b). Finally, the direct effect between the independent and dependent (path c) must be insignificant or substantially decrease while the indirect effect (path c’) remains significant (Preacher & Hayes 2008).

Hypothesis 3a was tested with a mediation model, presented with the following Figure, depicting all four conditional paths. Note that for these following two tests, the sample size was reduced to 98 (8 deleted) because the tests required answers to all the survey questions of each variable from all the participants, which was not the case for the preceding regressions.

Figure 3. Mediation Model with IV as Facebook Intensity

Note. N=96, *p <.05,**p<.01.

Firstly, it is noted that Facebook Intensity positively affects Social Capital (b=.2155, s.e=.0714, p=.0039), which was partly proven by H1c in the previous section. Even though Bonding showed no affiliation with Facebook Intensity, the Bridging portion accounts for enough change in this

Social Capital Facebook Intensity Self-Perceived Employability b=-.0564 c’ c b=.0748*

(32)

32 general Social Capital scale. Secondly, it was found that Social Capital positively related to Self-Perceived Employability (b=.3534, s.e=.1199, p=.004). The bootstrapping resampling method was then used with a 95% confidence interval (Hayes & Preacher, 2014). Proof of mediation was hence found with a significant path c’; the indirect effect between Facebook Intensity and Self-Perceived Employability (b= .0748, CI= .0279 to .1518), and so hypothesis H3a is accepted. Indeed, the correlation between LinkedIn Connections and Self-Perceived Employability is fully mediated by social capital.

The next model was identical to the previous, but testing for the dependent variable LinkedIn Connections instead;

Figure 4. Mediation Model with IV as LinkedIn Connections

Note. N=96, *p <.05,**p<.01.

Similarly, a significant affiliation was found between LinkedIn Connections and Social Capital, even though it was very low (b=.001, s.e= .0004, p=.0118). Path b was the same test as path b of Figure 3. While the direct effect of LinkedIn Connections on Self-Perceived Employability was insignificant (b=-.0004, s.e=.0005 p=.4523), the indirect was found to be very slightly positive and significant (b=.0004, CI= .0001 to .001), allowing H3b to be accepted. Therefore, it is proved that the effect of LinkedIn Connections on Self-Perceived Employability is completely mediated by Social Capital.

6. Discussion

This section will discuss and analyze the preceding supported and rejected hypotheses, as well as additional findings. Firstly, conclusions about Facebook will be drawn relating to social capital and employability. Secondly, LinkedIn will be discussed with similar structure. Subsequently, the results of data analysis of Social Capital and Self-Perceived Employability will be explored. This

b=-.0004 Social Capital LinkedIn Connections Self-Perceived Employability c’ b=.0004* c

Business school employability : to what extent do Social Network sites affect self-perceived employability through social capital?